44 KSME Journal, Vol. 3, No. 1, pp. 44~49, 1989.

UTILIZATION OF HONEYCOMB S T R U C T U R E S FOR A FLAT-PLATE

SOLAR COLLECTOR BY REDUCTION OF N A T U R A L
CONVECTION HEAT LOSS

Tae-Jin Kim* and Chong-Bo Kim**

(Received December 23, 1988)

A most common and economical way of utilizing solar energy is to use a flatplate collector to capture incoming solar energy
by heating either liquids or gases. Therefore, a solar collector should be the most critical part of the performance necessary for
the system, since maximum available heat depends solely on the collector. In the present investigation, considering that most heat
loss from solar collectors results from the natural convection between an absorber plate and a coverglass, it has been demonstrated
that this natural convection can be suppre~ed and heat performances of a solar collector are enhanced by placing thin and poorly
conducting honeycomb material between an absorber plate and a cowtrglass. By suppressing natural convection within collector
spacing it has been shown experimentally that honeycomb structures effectively raise critical Rayleigh number, since they provide
more shear surfaces.

Key Words : Honeycomb Structure, Solar Collector, Heat Use Ratio, Suppression

NOMENCLATURE : Solar collector tilt angle, deg

: Dynamic viscosity of air, k g / m sec
Ar : Absorber plate area, m
: Kinematic viscosity of air, m/sec
An : Aspect ratio, A n = H / L
: Density of air, k g / m
A~ : Horizontal aspect ratio, A w - W / L
: Absorptivity
c~ : Specific heat of air, K c a l/ k g ~
: Transmissivity
C, : Specific heat of water, K c a l/ k g ~
E : Heat use ratio, Q u s e / Q total
g : Gravity constant, 9.80665m/sec 2 1. INTRODUCTION
G~ : Grashof number, G r = g ' ~ ( T , , - T~)La/u 2
H : Length of honeycomb cell, m
: Instantaneous intensity of the solar radiation, Kcal/ The major heat losses from well-designed flat-plate solar
Hr
collectors are the result of conductive-convective-radiative
sec
heat transfer. The various methods; have been investigated in
K Conductivity of air, Kcal/m sec ~
order to reduce the heat losses from the solar collector for
L Height of honeycomb cell, m
many years. Hollands (1965a ; 1965b ; 1973) presented a sim-
m Mass flow rate of water, kg/sec
plified analysis of honeycomb solar collectors. Four aspects
Nu Nusselt number, Nu = UL 9 L / K
of honeycomb panels were discussed: natural convection,
Pr Prandtl number, Pr = IxCS K
thermal radiation and conduction heat transfers directly
Q loss Heat loss; quality, Kcal/m sec ~
through the honeycomb walls and finally the transmittance of
Q use Useful heat energy, Kcal/m sec ~
honeycomb to solar radiation. Buchberg, et al. (1971) indicat-
Q total Total heat energy, K c a l / m sec ~
ed that a rectangular cellular structure with the long side
Ra Rayleigh number, Ra =pagt~CpLa ZIT/#" K
running east-west was superior to fixed flat-plate solar col-
Ta Mean temperature of air, ~
lector because of higher solar transmittance to the absorber
7", Mean temperature of the inlet water, ~
plate. Arnold, et al. (1978) studied on the effects of aspect
T, Mean temperature of the absorber plate, ~
! ratio on liquidfilled rectangular honeycomb. The results
Thickness of honeycomb cell, m
showed heat loss Nusselt number was increased with higher
T~ Mean temperature of the coverglass, ~
tilt angles and higher aspect ratios 1W / H ) in the range of the
To Mean temperature of the outlet water, ~
Rayleigh numbers from 103 to l0 s. Cane, et al. (1977) experi-
U, Overall heat transfer coefficient, K c a l / m sec ~
mentally obtained Nusselt number-Rayleigh number results
1/2 Width of honeycomb cell, m
for the free convective heat-loss rate across honeycomb
panels heated from below and inclined with respect to the
Subscripts
horizontal. Meyer, et al. (1979) has studied the local and
: Volume coefficient of expansion for air, 1/~
average values of the Nusselt Number in moderate aspect
z/T : Temperature difference between the Tp and To, ~
ratio enclosures. The results show that aspect ratios between
*Iljin Electric and Machinery Co., Suwon 445-970, Korea 1 and 2 results in the highest heat transfer coefficients.
**Department of Mechanical Engineering, Inha University, Inchon The purpose of honeycomb is to suppress or dampen the
402-751, Korea free convection currents which would otherwise occur in the
 U T I L I Z A T I O N OF HONEYCOMB S T R U C T U R E S FOR A F L A T - P L A T E S O L A R COLLECTOR B Y REDUCTION OF... 45

air layer and increases the heat performance of solar collec-
tor. In the present analysis, the results have been extended to
determine optimum honeycomb sizes for a flat plate solar
collector by suppressing natural convection heat losses.
2. E X P E R I M E N T A L METHOD
Heat performances were measured and compared directly
from simultaneous installations of two solar collectors (Fig.
1), one with honeycomb structures fabricated from thin
polycarbonate sheet and the other without honeycomb struc-
tures. Tilt angles of 30, 45 and 60 deg. from the horizontal,
and honeycomb sizes(W x H) of 10 • 10mm, 10 x 20mm and
10 • 40mm were utilized. Fig. 2 shows honeycomb structures
utilized in the p r e ~ n t experimentation. Honeycomb cell
walls are made of transparent poly-carbonate, with a thermal
conductivity, 0.000795kcal/m s ~
Two Sam Sung collectors are placed on the top of the Inha
Engineering College building in Inchon, Korea (NL 37~ ',
L126"30'E).
A Pyranometer is placed on the top of the coverglass,
slightly above and paralled to the collector module surface, to
measure incoming solar radiation. Forty thermocouples (q~=
0.3mm, C type) are placed at various positions in the collector
to measure temperatures of the absorber plate, the coverg-
lass, the ambient, and the inlet and the outlet fluid. Signals
from sensing devices are sent to Data Logger (FLUKE,
2240C.), which consists of five scanner's.
The experiment began each day at about 10A.M. and
stopped at about 4P.M. during April-May, 1986. Experimental
data for typical days are plotted in Fig. 3-11. Most of the
days were cloudless or partly cloudy and the average wind
speed 2-3m/s. Mass flow rate supplied the solar collector was
set to a constant at 1.2g/m 2 rain.
Total heat loss from the solar collector is broken down as
following four areas : (1) Radiant heat loss between the two
faces of a honeycomb panel. (2) Conduction heat loss through
honeycomb walls. (3) Conduction heat loss through back and
side walls of a solar collector. (4) Natural convection heat
loss between an absorber plate and a coverglass. However, it
has been shown that the most important source of the total
heat loss in a flat plate solar collector is due to natural
conw~ction (Park, 1986).
For a solar collector, solar absorption rate, Q total, equals
the sum of the heat loss rate, O loss, ~,nd the heat use rate for
the load, Q use ;

Q total = Q use + Q loss (1)

The solar absorption rate can also be expressed as the

product of glazing transmittance, r, plate absorptance, a,
plate area, Ar and the instantaneous intensity of solar radia-
Fig. I Photograph of experimental set-up tion incident on glazing, H r ;

HONEYCOMB STRUCTURES Q total = A ~ H r rA (2)

Heat loss rate can be written in terms of an effective

overall heat transfer coefficient, Ut, plate area, Ac, and the
temperature difference between average absorber plate and
ambient ( T ; - T~) ;

Q l o s s = A t UL(T~-- To) (3)

The heat use rate for the load can also be written in terms
of mass extraction rate, rh, specific heat of water, Cw, and the
temperature difference between average outlet water and
inlet water ( T o - T~) ;
ABSO]2F~ERPLA/ETE~ . j j CO~ER GLASS
Q use = mC~( T o - 7",) (4)

Therefore the heat use ratio (E) can be written as follows ;

Q use _ m C w ( T o - Ti) (5)

NO T~PF { L I It W t R~=}VLAw=W/L E- Q total A,:Hr ra
I SO(]ARE 28 10 1 10 0 .5 0.3570.357
CE L L can _r_,~ m n~n ,
RECTAN 28 20 ~5 10.7740 .3=17 3. R E S U L T S A N D D I S C U S S I O N

CELL } 28 - 40 I
Figures 3 - 1 1 show experimental results with or without
honeycomb structures in the collectors. Measured tempera-
Fig. 2 Honeycomb structures in a solar collector ture difference between outlet and inlet of the collector, and
 46 Tae-Jin K i m and Chong-Bo Kirn

21 *,ZOO .~ IZO0
9 ; KI-.O.CTCP*e ; Np-apm[vcone
me~(rcome 9 : NONs
Rnel{Nl ~[~e. --: RMeIs I[RP.
18
1000 iOOO
C u
~ 45 %
15
800 ~
02
g z
ua
3O 600 ~-
~s e-
8
z
. . . . . . . .
400 Z ku r z
6
~: 15"
(z; ~c
200 200
3

0 0 0 0 0 0
IO 11 12 13 l& 15 16 10 11 12 13 14 15 16 17
TIME [HR) "rIME IHR)

Fig. 3 Temperature difference and insolation Fig. 6 Temperature difference and insolation
10 • 10 honeycomb and no-honeycomb at 30 deg 10x 20 honeycomband no-honeycombat 30 deg

~200
9 t20O 21 ~ : Np-i~lps
21t . ! NP-~ON[ICO.. H~Ws
9 HONCTC~I

lgllL~tT/~H [~/Mx I 10D0

u
9 IOOO
c
45
15 15}
BOO ~ 8O0
s
ua
- g LU
9 600 ~- ~- 3C 600
oc
d
co
I 400 z W 400 Z
6 ~
1 200
200
3 ~

t 0 I 0
0 0
]0 11 12 13 14 IS 16 17
10 11 12 t~ l& 15 IS 17
TIME {HR} TIME ~HR]

Fig. 4 Temperature difference and insolation Fig. 7 Temperature difference and insolation
10 • 10 honeycomb and no-honeycomb at 45 deg 10 • 20 honeycomb and no.honeycomb at 45 deg

t2 O0
~N~rcone

~j
t --' JtJS~Lwr/~N {w/i,]~)
18--~ P4agls I[RP.
I000 1000

&5 15 45 15

800 :~ ~00 ~

~3oS 600 ~ i.oo

m
400 z 4~ :~

200 200

d 0 l0 0 6
9 10 II 12 I3 14 15 16 17 l0 II IZ 13 14 15 16
TIME [HR3 TIME 1MR]

Fig. 5 Temperature difference and insolation Fig. 8 Temperature difference and insolation
10 x 10 honeycomb and no-honeycomb at 60 deg 10 • 20 honeycomb and no-honeycomb at 60 deg

the ambient temperature are plotted with insolation during
the day. Various honeycomb sizes and various tilt-angles
were utilized for the performance comparisions for these two
installations. It is found that the temperature difference of the
collector performance is basically proportional to insolation9
The free convection heat loss between the absorber plate
and the coverglass may be explained to be due to the loss
from the Bernard cell flow pattern expected in the top-heavy
type instability driven by the component of gravity normal to
the heated bounding surface, and the loss caused by the
parallel flow motion driven by the component of gravity
along the heated bounding surface of the honeycomb panel.
The third source of the heat loss is believed to be due to the
secondary motion from the stratification of the flow in the
side wall regions. Each of these flow patterns is expected to
interact each other to a greater or less extent with various
honeycomb sizes and tilt-angles.
It is shown that the temperature difference between the
inlet and outlet water temperatures of the collector with
honeycomb structure are generally larger than that without
honeycombs. These results show honeycomb structures suc-
cessfully suppress the natural convection flow pattern in the
 U T I L I Z A T I O N OF HONEYCOMB S T R U C T U R E S FOR A F L A T - P L A T E SOLAR COLLECTOR B Y REDUCTION OF,,. 47

I~00 1
- - : IIl~lll[l~l 1El#'.
IOOO
6S
ac
.9
O00 : i
e-
~J
e- ~0
.8
600 ~,- [] 0

0
.7
~/: IS:

o.6
r
0 ,0 A
10 ]I 12 13 16 IS 16 17
TIME (MR}

Fig. 9 Temperature difference and insolation

10 • 40 honeycomb and no-honeycomb at 30 deg

O : 30 DEG.

]iI ........
~1 iA : HO-flJll(TCO,~e IEOO

IOOO
.3
[3 :
:
45
6(]
DEG.
DEG,

% .2
800

12
.I
600
cr

#00 0
G
~: ~s 103 104 10
200 RAYLEIGH NUMBER

Fig. 12 Experimental results in terms of heat use ratio versus

0 0 Rayleigh number for 10 x ].0mm honeycomb
L0 11 12 13 16 IS 16 17
TIME [HR)

Fig. I0 Temperature difference and insolation

10 • 40 honeycomb and no-honeycomb at 45 deg
1.
A ; IdP'-R~(TClmO - - E00

1
_.: pH,T4gjEI4TT[I%". .9
--: JR$1~LAT|IIp IVIH~) aooo

15
.8

60g ~ ,7
lz:
Cs
ul
100
o.6
3 200

a ~J
lO II 12 13 Ik IS IB
TIME IH~] E-*

Fig. Ii Temperature difference and insolation

10 • 40 honeycomb and no-honeycomb at 60 deg
O : 30 DEG.
.]
L'I : 45 DEG.
4 : 60 DEG.
collector spacing, resulting in reduction of heat loss. How-
.2
ever, as shown in Fig. 10~Fig. 11, temperature difference of
the collector with honeycomb of 10• show smaller
values than those without honeycomb 9 .I
Figures 12-15 represent the heat use ratio-Ra plots for
various tilt-angles. The longitudinal motion is expected to
play an important role to the heat transfer.
10 3 10" 0s
The results of all figures present indirectly that the flow
RAYLEZGH NVMBER
induced by the topheavy instability tends to generate the
longitudinal roll and that it is a major factor of heat loss. Fig. 13 Experimental results in terms of heat use ratio versus
Therefore, the effect of suppressing the natural convection at Rayleigh number for 10 • 20mm honeycomb
 48 Tae-Jin Kim and Chong-Bo K i m

10

.9 ?
GO ~
Oo

.8 8 0 o
0 0

7
OE1 O O
O
B%o
.7 n
o cfh
o .6 ~6 %1
At,
r.&

m.5 ~5

,<

:r:

O : 30 DEG. O : I0 x 10 HONEYCOMB
.3 ?
O : 45 DEG. O : 10 x 2 g HONEYCOMB
& : 60 DEG. : ]0 x 4 0 HONEYCOMB
-- : NO - HONEYCOMB
.2 2

.1 1

0
10 3 10 4 10 10 3 10 4 10 5
RAYLE IGH NUMBER RAYLEIGH NUMBER

Fig. 14 Experimental results in terms of heat use ratio versus Fig. 16 Experimental results in terms of heat use ratio versus
Rayleigh number for 10 x 40mm honeycomb Rayleigh number at 30"

10

.9

.8 0 OO
O

.7
0 0
oc~] %00 o %
no O0
E]~ r-1
o 6
o6

~5 %a ~a
5 t'2t
A
E~
< m~
:z:

O : 10 x ]0 HONEYCOMB
O : 30 DEG.
[] : l0 X 2 0 HONEYCOMB
: 45 DEG. : I0 • HONEYCOMB
~, : 60 DEG.
-- : NO - HONEYCOMB

10 3 10 4 10 s
10 3 10 4 10 s

RAYLEIGH NUMBFR RAYLEIGH NUMBER

Fig. 15 Experimental results in terms of heat use ratio versus Fig. 17 Experimental results in terms of heat use ratio versus
Rayleigh number for no-honeycomb Rayleigh number at 45"
 U T I L I Z A T I O N OF HONEYCOMB S T R U C T U R E S FOR A F L A T - P L A T E S O L A R COLLECTOR B Y REDUCTION OF... 49

strated that the suppression of this free convection is possible

by installing honeycomb structures in the collector spacing.
.9 Employing two collectors (one with honeycombs and the
other without), the results were compared simultaneously for
their performances. The heat use ratio for the collector with
.8 10 • 10mm honeycomb increased approximately 29.5% at the
collector tilt angle 0=30 ~ 18.5% at 0=45 ~ and 25.3% at 0 =
60~ compared with that of the collector without honeycombs.
.7 O0 0 The honeycomb structures are appeared to be more effective
% CPo at low tilt angles. At a fixed tilt angle 0=30 ~ the heat use
o,6 ~ 1 7 6 1 7 6~
1 7o6 [] ~ 0% ratio increased approximately 29.5% in 10x10mm honey-
O9[3 O0 o comb, 17.1% in 10• but decreased 0.7% in 10•
t~
honeycomb.
.5
u) The results of the present investigation may be limited to
Inch 9 Korea (NL 37~ ', L126~ However, the data
show consistency with various incoming solar radiations and
different tilt angles. Therefore, introducing honeycomb struc-
0 : 10XI0 HONEYCOMB tures into collector spacings can be effective means to
.3 [] : 10X~ HONEYCOMB improve heat performances of flat-plate solar collectors.
: 10X40 HONEYCOMB
Especially, at tilt angle of 30", it is noted that the heat
-- : NO - HONEYCOMB
.2 utilization rate of the solar collector for 10• honey-
comb has been improved approximately 30% which is signifi-
cant in any practical application.
.1

10 3 10 4
RAYLEIGH NUMBER
Fig. 18 Experimental Resultsin terms of Heat Use Ratio versus
Rayleigh number at 60~
0=30 ~ is superior to that of 0=45 ~ and 0=60 ~ For the
collector without honeycombs, the heat-use-ratio is found to
be superior at 0=45 ~ compared with the values of 8=30 " and
0=60 . .
Figures 16 and Fig. 18 show the heat use ratio as a function
of Ra.
Figures 16 shows the effects of the suppression from free
convection currents with various honeycomb sizes at 0 = 30~
The suppression employing honeycomb cells is found to be
effective means to increase collector heat efficiencies. With
honeycomb sizes of 10 • 10ram and 10 x 20mm, the results give
considerable improvement on the performance of the collec-
tors.
4. CONCLUSIONS
The heat loss in a solar collector results mainly from the
natural convection between an absorber plate and a coverg-
lass. Therefore, it is important to suppress this free convec-
tion heat loss in order to increase heat performances of a
collector.
In the present investigation, it has been successfully demon-
REFERENCE
I0 S
Arnold, J.N., Edwards, D.K., 1978, "Effect of Cell Size on
Natural Convection in high L/D, Tilted Rectangular Cells
Heated and Cooled on Opposite Faces", ASME Paper No. 18
-WA/HT-5.
Buchberg. H., Lalude, O.A., and Edwars, D.K., 1971, "Per-
formance characteristic of Recangular Honeycomb Solar-
Thermal Convecters", Solar Energy, Vol. 13, pp. 213-221.
Cane, R.L., Holland, K.G.T., Raithby, G.D., and Unny, T.E.,
1977, "Free Convection Heat Transfer across Inclined Honey-
comb Panels", ASME J. of Heat Transfer, Vol. 99, pp. 86-91.
Charters, W.W.S., and Peterson, L.F., 1972, "Free Convec-
tion Suppression Using Honeycomb Cellular Materials",
Solar Energy, Vol. 13, pp. 353-361.
Hollands, K.G.T., 1965a, "The Use of Honeycomb Devices
in Flat Plate Solar Collectors", Solar Energy, Vol. 9, No. 3,
pp. 159-164.
Hollands, K.G.T., 1965b, "Honeycomb Devices in Flat Plate
Solar Collector", Solar Energy, Vol. 9, No. 3, pp. 554--558.
Hollands, K.G.T., 1973, "Natural Convection in Horizontal
Thin-Walled Honeycomb Panels", J. of Heat Transfer,
ASME, series C, Vol. 95, pp. 439-444.
Meyer, B.A., Mitchell, J.W., and El-Wakil, M.M., 1979,
"Natural Convection Heat Transfer in Moderate Aspect
Ratio Enclosures", ASME J. of Heat Transfer, Vol. 101, pp.
655-659.
Park, S.Y., 1986, "The Influence of Honeycomb Sizes on
Solar Collector Performance", Master's Thesis, Inha Univer-
sity.